yzjdz

ValueError: Dimensions must be equal, but are 256 and 384 for 'tower_0/build_pyramid/build_P4/add' (op: 'Add') with input shapes: [1,?,?,256], [1,?,?,384].

3

I try to training my dataset, and i have reference the three steps in README.md, I use the pretrained weights resnet_v1_50.ckpt, and i modify the $PATH_ROOT/libs/configs/cfgs.py and $PATH_ROOT/libs/label_name_dict/lable_dict.py and $PATH_ROOT/data/io/read_tfrecord.py,...

In the COCO_Caption_Scorer calculation process(class ImageCaptionDataset), jieba was not used for word segmentation; Instead, labels and predictions were directly used for metric calculation. Is this reasonable? I used VLMEvalKit to...